Objective: To determine the effects of walking meditation (WM) on functional performance, disease severity, and anxiety in Parkinson's disease (PD). Design: This was a randomized controlled trial. Settings: The study was conducted at a regional hospital. Subjects/Interventions: Thirty-three participants with PD were randomly allocated to the control (CON) group (n = 16) or the WM group (n = 17). Participants in the WM group were asked to perform WM monthly under supervision and encouraged to practice at home at least 3 days/week for 12 weeks. Outcome measurements: Gait velocity, Timed Up and Go, five times sit to stand (FTSTS) test, Unified Parkinson's Disease Rating Scale (UPDRS), and the percentage of participants with anxiety (Hospital Anxiety and Depression Scale-part anxiety [HADS-A] ≥8). Results: Both groups showed reduced gait velocity (p < 0.05), although impairment of the FTSTS (p < 0.05) score was observed only in the CON group. A significant enhancement within and between groups in the total UPDRS and UPDRS part II scores was observed only in the WM group. The percentage of participants with anxiety (HADS-A ≥ 8) decreased significantly only in the WM group (p < 0.05), compared with the baseline and after 12 weeks. There was no loss to follow-up in the WM group, and the participation rate of training was 3.2 days/week. Conclusions: Home-based WM can encourage high rates of exercise adherence, reduce disease severity, lower the percentage of participants with anxiety, and might be suitable during disease endemic and/or pandemic in PD. The protocol was registered on Thaiclinicaltrials.org (Identifier: TCTR20201009001).